The final part of this blog on cholangiocarcinoma is of course the surgery. I personally find surgery for hilar cholangiocarcinoma technically challenging. There is no doubt that it is some of the technically most complex and difficult surgery that we do. The major technical problems with hilar cholangiocarcinomas can be divided into
1. having a sufficient future liver remnant volume
2. obtaining clearance of the vessels that need to be preserved to supply the future liver remnant
3. obtaining clear ductal margins
4. adequate reconstruction for biliary drainage.
Vessel clearance- The mantra of liver surgeons is that safe liver surgery demands a minimum of two contiguous segments of liver with arterial and portal inflow and hepatic venous outflow. In practical terms when it comes to resecting hilar cholangiocarcinoma this can be difficult and tumours are often infiltrative involving the portal vein confluence or hepatic arteries. From the perspective of the portal vein, involvement of the portal vein is often amenable to resection and reconstruction and this can be done with either a patch, an end to end anastomosis or a tube graft.
The two cases below are both cholangiocarcinomas with a mass forming component and both had involvement of the portal veins.
The first case had complete occlusion of the right anterior sectoral branch and 90% occlusion of the left portal vein. The right posterior sectoral branch of the right portal vein was patent but because of the position of the tumour and involvement of the right hepatic artery this patient could not undergo extended left hepatectomy. Instead an extended right hepatectomy was performed with end to end reconstruction of the left portal vein.
The second case is a patient with a large intrahepatic cholangiocarcinoma which also involved the left portal vein.
This was resected and reconstructed using the same principle which exploits the long length of the left portal vein which often leaves redundant length in the “backward 7” conformation as I like to call it.
I like this technique because it allows you to stay far away from the tumour itself. I usually do the initial portal dissection and establish that the artery is clear and that the portal vein is reconstructable. In doing so a thorough lymphadenectomy is also performed. I then do the vascular reconstruction first the logic being that if this is not possible then the rest is not going to happen. Using fine vascular clamps to clamp the left portal vein and main portal vein I divide the vein in two places to the left and below the tumour. There may be some back bleeding which can be controlled either with a bulldog or fine clamp prior to closure using 6/0 prolene. The portal vein reconstruction I do using continuous 6/0 prolene. I do not tie the suture however until the clamps have been released and I am completely happy that there is no narrowing of the vein lumen. Not tying vascular anastomotic running sutures is always alarming to people who have not seen this before but there should be little or no bleeding if the anastomosis has been done well and allowing the vein to stretch up is important. If the vein has not fully stretched up you can tie the suture loose to leave a growth factor to allow further expansion. The left portal vein reconstruction is prone to kinking particularly if it is too long. If it looks like it is kinked my advice is to re-clamp and redo the anastomosis resecting more of the vein. It is also important to reconstruct the falciform and possibly also left triangular ligament of you are doing a right trisegmentectomy (extended right hepatectomy) to prevent rotation or movement of the future liver remnant.
Here is an early CT scan from the first patient showing the reconstructed left portal vein and you can see the patient has has a right trisegmentectomy.
Arterial involvement to my mind usually means that the patient has non-resectable cancer. The problem is not so much the ability to reconstruct the artery although this can be the case but more that it reflects poor prognosis and the outcome of patients who have arterial resection is often very poor. Other people are more aggressive and would consider arterial resection routinely however survival figures often do not support such aggressive surgery and patients who are well palliated in terms of their jaundice may have better overall median survival than those undergoing resection requiring arterial reconstruction.
Obtaining clear ductal margins
Cholangiocarcinoma is a difficult disease and is prone to spreading along the ducts. It is often difficult to be certain where the duct is clear and there may be anatomic restrictions which determine where a duct must be divided. For this reason in many series microscopic positive margins are not uncommon. International Classifications consider tumour within 1 mm of the resection margin as also being positive. Some centres notably in Japan perform routine frozen section analysis of bile duct margins. The pathologists doing this must be highly skilled and specialised as the interpretation of frozen section material can be quite difficult. More important is probably good quality imaging and good pre-operative planning to achieve clear margins and avoid operating on patients who are certain to have positive margins.
The other factor which is sometimes overlooked is that the caudate lobe is almost invariably invovled in hilar cholangiocarcinoma and it should in my opinion be removed every time in central cases and in any case where there is radiological or intraoperative ulatrasound evidence of caudate lobe biliary dilatation.
The techniques for resection of hilar cholangiocarcinoma are described very well in the published literature and so I will not go into this further. These papers below from Peter Neuhaus group in Germany are good and of course the greatest technical expertise in resecting hilar cholangiocarcinoma is in Japan. Yuji Nimura in particular explains very well the evidence base behind different approaches to hilar cholangiocarcinoma.
Here is a nice picture of Yuji Nimura (Nagoya) and Thomas van Gulik (Ansterdam) which I took in Belgrade at a meeting recently. On the right is Yuji’s drawing of the hilar plate and the anatomical arrangements of the ducts.
Oncological superiority of hilar en bloc resection for the treatment of hilar cholangiocarcinoma. Neuhaus P, Thelen A, Jonas S, Puhl G, Denecke T, Veltzke-Schlieker W, Seehofer D.Ann Surg Oncol. 2012 May;19(5):1602-8.
Extended resections for hilar cholangiocarcinoma. Neuhaus P, Jonas S, Bechstein WO, Lohmann R, Radke C, Kling N, Wex C, Lobeck H, Hintze R. Ann Surg. 1999 Dec;230(6):808-18; discussion 819.